Projecting the ACC's 2014 conference race using advanced stats

Our general perception of the ACC title game is that an undefeated Florida State team is going to take on some 4-4 Coastal Division champion. But FSU still has some work to do to ensure the former, and Duke is threatening to completely ruin the latter. Let's see what the numbers say.

Win projections

Below are the projected ACC win totals, based on win probabilities culled from the current F/+ ratings. These ratings are based on full-season performance with no personnel or momentum adjustments, so all proper disclaimers -- FSU was without suspended quarterback Jameis Winston for the Clemson game, Clemson is without injured quarterback Deshaun Watson for a few weeks, Virginia Tech's rating has been boosted by Ohio State's recent surge, et cetera -- apply.

Odds of ACC conference records

The distribution of teams' rankings above is pretty impressive. The top three teams in the conference are all in the Atlantic division, and six of seven teams in the Coastal are within 14 spots of each other (20th to 33rd). Poor Clemson, Louisville, and Boston College would all be serious contenders in the Coastal.

Wait, so Florida State is going to lose?

ATLANTIC
Major contenders: Florida State, Clemson
Minor contenders: none

Well ... maybe. FSU has crept back into the F/+ top 10 and is still getting slight dings for Winston missing the Clemson game and for Oklahoma State's J.W. Walsh getting injured right after the FSU game. So all of the Seminoles' win probabilities are probably slightly low because of that.

Still, they've survived unscathed. They held off Clemson and Notre Dame in Tallahassee, and they pulled past Oklahoma State and N.C. State away from home. As I've written a few times already, they've mastered the art of playing well when they need to (and often not a moment sooner). And as their ranking improves, it starts to feel safer to assume that they will continue to figure things out and begin to peak in November, a la Auburn in 2010 or 2013.

A lot of our assumptions, however, are premised around the helmet FSU wears. If Clemson were 7-0 with a few close wins, we wouldn't trust that the Tigers would continue to win. If Louisville or BC were in this position, we really wouldn't trust it. Because FSU won the national title last year, and because we recognize so many star-caliber names on the two-deep, we perhaps justifiably assume the 'Noles will be just fine. I know I do.

But the numbers aren't programmed to see that. They only see a team that has played the way the 'Noles have. They say this team might only have about a 58 percent chance of winning on the road against a Louisville team with one of the nation's toughest, most aggressive defenses. They see a team with only about a 68 percent chance of beating a Miami team that can move the ball. And they see a team that has been rickety enough at home that, while they have good odds of beating teams like Virginia and BC, those odds are only in the 82 percent range, not the 98 percent range.

Conservative odds being what they are, then, there's still a decent chance that FSU loses a game. I assume that won't happen, and so do you, but it's not off the table.

Operation 4-4

COASTAL
Serious contenders: Duke, Pittsburgh
Minor contenders: Virginia Tech (and maybe Virginia or Georgia Tech)

One of our favorite memes this season (well, most recent seasons) is the thought of an all-directions 4-4 tie in the Coastal Division. The Coastal has already formed a perfect transitive circle, which feeds the idea that anybody can beat anybody. But Duke's positioning -- 2-1 with wins over Virginia and Georgia Tech and a top-25 overall ranking -- threatens to kill that idea before it gets too far.

The Devils are given about a 96 percent chance of finishing 5-3 or better, which means most dreams of a seven-way 4-4 tie die on the vine. But they did lose at Miami already, and after a Week 9 bye week, they do get Pitt and Syracuse on the road. Lose to fellow division favorite Pitt, and things get interesting. Lose in an upset in the Carrier Dome, and Project Slog's odds of success increase dramatically.

For now, though, Duke and Pitt have a clear leg up on everybody else. That makes their Week 10 battle at Heinz Field pretty big.

Sure bets

Here are the remaining in-conference win probabilities, broken into ranges.

70+% games: Duke 3, Virginia Tech 3, Boston College 2, Clemson 2, Florida State 2, Pittsburgh 2, Georgia Tech 1, Miami 1, N.C. State 1, Virginia 1

Because Syracuse stinks, and because Virginia Tech and UNC have to come to Durham, Duke easily has the best odds of winning out among the Coastal's 2-1 teams. Meanwhile, for Project Slog to come to fruition, UNC is going to have to maintain the improvement that allowed it to get past Georgia Tech last week. The Heels might need to figure out how to play at least a little bit of defense, too. That's probably too much to ask.

Slog Level Delta

Here are the odds of a division champion finishing with a particular record:

0.0 percent on the 4-4 line. Sorry.

This Week

• Miami at Virginia Tech (Thursday, 8:00 p.m. ET, ESPN). Win probability: Virginia Tech 70%.
• North Carolina at Virginia (Saturday, 12:00 p.m. ET, ESPN3). Win probability: Virginia 91%.
• Boston College at Wake Forest (Saturday, 3:30 p.m. ET, ESPN3). Win probability: BC 94%.
• Georgia Tech at Pittsburgh (Saturday, 3:30 p.m. ET, ESPNU). Win probability: Pitt 59%.
• Syracuse at No. 21 Clemson (Saturday, 7:00 p.m. ET, ESPNU). Win probability: Clemson 99%.

It's hard to imagine Virginia having a 91 percent chance of beating North Carolina, especially after the Heels nearly beat Notre Dame in South Bend and took down Georgia Tech last week. It's possible that UNC's ratings are taking too much of a hit from earlier performances -- here's your reminder that the Heels allowed 70 points to East Carolina, 50 to Clemson, and 29 to Liberty in September -- and that Virginia is getting too much credit for slowing down a UCLA offense that has been great since Week 1. Regardless, Virginia should probably be considered more of a favorite than it is in your head.

Meanwhile, the most important game of the week is Georgia Tech-Pitt, as everybody else's Coastal Division hopes require both Pitt and Duke to lose as a favorite once or twice.

All projections